The availability of the human and mouse genome sequences has allowed the identification and comparison of their respective degradomes—the complete repertoire of proteases that are produced by these organisms. Because of the essential roles of proteolytic enzymes in the control of cell behaviour, survival and death, degradome analysis provides a useful framework for the global exploration of these protease-mediated functions in normal and pathological conditions.
One such family of proteases is the C19 hydrolases. The ubiquitinyl hydrolases of family C19 are intracellular peptidases that remove ubiquitin molecules from polyubiquinated peptides by cleavage of isopeptide bonds. The purpose of the de-ubiquitination is thought to be editing of the ubiquitin conjugates, which could rescue them from degradation, as well as recycling of the ubiquitin. The ubiquitin/proteasome system is responsible for most protein turnover in the mammalian cell, and with over 50 members, family C19 is one of the largest families of peptidases in the human genome.
The peptidases of family C19 have a more complicated structure than the ubiquitinyl hydrolases in the related family C12. Many of the proteins have multiple domains. The peptidase unit contains a ‘fingers domain’ that is a four-stranded beta sheet that interacts with the ubiquinated substrate. The USP proteolytic system in human tissues is extremely complex.
The Ubiquitin Pathway
Ubiquitin-mediated proteolysis is an important pathway of non-lysosomal protein degradation which controls the timed destruction of many cellular regulatory proteins including, p27, p53, p300, cyclins, E2F, STAT-1, c-Myc, c-Jun, EGF receptor, IκBα, NFκB and β-catenin (reviewed in Pagano, 1997, FASEB J. 11: 1067). Ubiquitin is an evolutionary highly conserved 76-amino acid polypeptide which is abundantly present in all eukaryotic cells. The ubiquitin pathway leads to the covalent attachment of a poly-ubiquitin chain to target substrates which are then degraded by the multi-catalytic proteasome complex (see Pagano, supra, for a recent review). Many of the steps regulating protein ubiquitination are known. Initially the ubiquitin activating enzyme (E1), forms a high energy thioester with ubiquitin which is, in turn, transferred to a reactive cysteine residue of one of many ubiquitin conjugating enzymes (Ubcs or E2s). The final transfer of ubiquitin to an e-amino group of a reactive lysine residue in the target protein occurs in a reaction that may or may not require an ubiquitin ligase (E3) protein. The large number of ubiquitin ligases ensures a high level of substrate specificity.
The Ubiquitin Pathway and the Regulation of the G1 Phase by F Box Proteins
Genetic and biochemical studies in several organisms have shown that the G1 phase of the cell cycle is regulated by the ubiquitin pathway. Proteolysis of cyclins, Ckis and other G1 regulatory proteins is controlled in yeast by the ubiquitin conjugating enzyme Ubc3 (also called Cdc34) and by an E3 ubiquitin ligase formed by three subunits: Cdc53, Skp1 and one of many F box proteins (reviewed in Patton, et al., 1998, Trends in Genet. 14:6). The F box proteins (FBPs) are so called because they contain a motif, the F Box, that was first identified in Cyclin F, and that is necessary for FBP interaction with Skp1 (Bai, et al., 1996, Cell 86:263). Cdc53 (also called Cul A) and Skp1 appear to participate in the formation of at least three distinct E3s, each containing a different FBP. Because these ligases are similar protein modules composed of Skp1, Cul A, and an FBP, they have been named SCF. The three SCFs identified so far in S. cerevisiae are: SCFCdc4 (which recruits the Ckis Sic1 and Far1, the replication factor Cdc6, and the transcriptional activator Gcn4, as substrates through the F-Box protein Cdc4), SCGrr1 (which recruits the G1 cyclins Cln1 and Cln2 as substrates through the F-Box protein GRR1), and SCFMet30 (which recruits the G1 cyclin Cln3 as a substrate throughout the F box protein MET30; see Pagano and Patton, supra, for recent reviews).
The interaction of SCF ligase with its substrates occurs via the FBP. FBPs are present in all eukaryotes (at least 54 in mammals; Cenciarelli, et al., 1999, Current Biol. 9: 1177; Winston, et al., 1999, Current Biol. 9: 1180). In addition to the F Box, many FBPs contain additional domains that facilitate both protein:protein interactions, e.g. WD-40 domains or leucine-rich repeats (LRRs), and protein:DNA interactions, e.g. tankyrase binding domains or HNH domains. Since the substrate specificity of SCF ligases is dictated by different FBPs that act as substrate targeting subunits, the large numbers of FBPs with varying combinations of protein or DNA interaction domains ensure highly specific substrate recognition (Cenciarelli, et al., supra; Winston, et al., supra).
FBP1, A Mammalian FBP Involved in Regulation of APC/C
Fbp1, the mammalian homolog of Xenopus β-TrCP1 (β-transducin repeat containing protein) (Spevak, et al., 1993, Mol. Cell. Biol. 8:4953), was identified using Skp1 as a bait in a two-hybrid screen (Cenciarelli, et al., supra). Fbp1 is an F box protein containing seven WD-40 domains (Margottin, et al., 1998, Mol. Cell. 1:565), and is involved in the degradation of IKBα. family members in response to NFKB activating stimuli (Gonen, et al., 1999, J. Biol. Chem. 274:14823; Hatakeyama, et al., 1999, Proc. Natl. Acad. Sci. USA 96:3859; Hattori, et al., 1999, J. Biol. Chem. 274:29641; Kroll, et al., 1999, J. Biol. Chem. 274:7941; Ohta, et al., 1999, Mol. Cell. 3:535; Shirane, et al., 1999, J. Biol. Chem. 274:28169; Spencer, et al., 1999, Genes Dev. 13:284; Winston, et al., 1999, Genes Dev. 13:270; Wu and Ghosh, 1999, J. Biol. Chem. 274:29591; Yaron, et al., 1998, Nature 396:590). In addition, consistent with the finding that Xenopus and Drosophila Fbp1 orthologs act as negative regulators of the Wnt/β-catenin signaling pathway (Jiang and Struhl, 1998, Nature 391:493; Marikawa and Elinson, 1998, Mech. Dev. 77:75), several studies report that human Fbp1 controls .beta.-catenin stability in vitro and in mammalian cultured cells (Ilart, et al., 1999, Curr. Biol. 9:207; Ilatakeyama, et al., supra; Kitagawa, et al., 1999, EMBO J. 18:2401; Latres, et al., 1999, Oncogene 18:849; Winston, et al., 1999, Genes Dev. 13:270).
To date, well-characterized substrates of mammalian Fbp1 have been found to share a common destruction motif, DSGxxS, and are recognized by Fbp1 only upon phosphorylation of the two serine residues present in this motif. There is, however, some recent evidence for additional mammalian substrates of Fbp1 lacking a completely conserved binding domain, such as ATF4 (Lassot, et al., 2001, Mol. Cell. Biol. 21:2192), Smad3 (Fukuchi, et al., 2001, Mol. Biol. Cell 12-1431), NFKB p105 (Orian, et al., 2000, EMBO J. 19:2580) and NFKB p100 (Fong and Sun, 2002, J. Biol. Chem. 277:22111). A conserved DSGxxS motif is present not only in Fbp1 substrates but also in certain regulators of Fbp1, such as the HIV protein Vpu, which targets Fbp1 to the non-physiological substrate, CD4, in virally infected cells. (Margottin, et al., supra). The DSGxxS destruction motif may also be found in peptide regulators of Fbp1 termed pseudosubstrates; however, pseudosubstrates escape the normal degradation fate of other FBP target proteins and instead modulate the activity of the FBP, and corresponding Cks, such as cellular localization and substrate targeting. For example, the Fbp1 pseudostibstrate hnRNP-U not only inhibits Fpb1 from targeting inappropriate substrates but also serves to localize Fbp1 to the nucleus (Davis, et al., 2002, Genes Dev. 16:439).
A further level of complexity is added by the presence of a Fbp1/β-TrCP1 paralogous gene product, called β-TrCP2 or Fbxw1B (78% identical, 86% similar; Kipreos and Pagano, 2000, Genome Biology 1:3002.1). Fbp1 and β-TrCP2 are ubiquitously expressed in adult human tissues (Cenciarelli, et al., supra; Koike, et al., 2000, Biochem. Biophys. Res. Commun. 269:103). In addition, β-TrCP2 has biochemical properties similar to Fbp1 in its ability to sustain the ubiquitinylation of both β-catenin and IKBα. family members in vitro and to control their degradation in mammalian cultured cells (Fuchs, et al., 1999, Oncogene 18:2039; Suzuki, et al., 1999, Biochem. Biophys. Res. Commun. 256:127; Tan, et al., 1999, Mol. Cell. 3:527). Despite these similarities, Fbp1 localizes to the nucleus and β-TrCP2 mainly to the cytoplasm (Davis, et al., 2002, Genes Dev. 16:439). It is not clear whether these two FBPs have overlapping functions in vivo, or if each of them recognizes specific substrates.
Deregulation of the Ubiquitin Pathway in Cancer and Other Proliferative Disorders
Cancer develops when cells multiply too quickly. Cell proliferation is determined by the net balance of positive and negative signals. When positive signals overcome or when negative signals are absent, the cells multiply too quickly and cancer develops.
Ordinarily cells precisely control the amount of any given protein and eliminate the excess or any unwanted protein. To do so, the cell ubiquitinates the undesired protein to tag the protein for proteasome degradation. This mechanism goes awry in tumors, leading to the excessive accumulation of positive signals (oncogenic proteins), or resulting in the abnormal degradation of negative regulators (tumor suppressor proteins). Thus, without tumor suppressor proteins or in the presence of too much of an oncogenic protein, cells multiply without control, forming tumors (reviewed by Ciechanover, 1998, EMBO J. 17: 7151; Spataro, 1998, Br. J. Cancer 77: 448). For example, abnormal ubiquitin-mediated degradation of the p53 tumor suppressor (reviewed by Brown and Pagano, 1997, Biochim. Biophys. Acta 1332:1), the putative oncogene β-catenin (reviewed by Peifer, 1997, Science 275:1752) and the Cki p27 (reviewed in Ciechanover, supra; Spataro, supra; Lloyd, 1999, Am. J. Pathol. 154: 313) have been correlated with tumorgenesis, opening to the hypothesis that some genes encoding ubiquitinating enzymes may be mutated in tumors.
Initial evidence indicates that human F box proteins play a role in the ubiquitination of G1 regulatory proteins as do their lioniologues in yeast. Unchecked degradation of cell cycle regulatory proteins has been observed in certain tumors and it is possible that deregulated ubiquitin ligase plays a role in the altered degradation of cell cycle regulators. A well understood example is that of Mdm2, a ubiquitin ligase whose overexpression induces low levels of its substrate, the tumor suppressor p53.
Alternately, F box proteins have been shown to interact directly with DNA regulating proteins or DNA itself (see below). F box proteins in yeast are known to regulate genomic stability and senescence, and recent data has shown that F box inhibition in mammalian cells can lead to the loss of DNA damage checkpoints. The identification of novel F box protein substrates or activity may thus extend the role of F box proteins in tumorigenesis beyond the understood regulation of traditional cell cycle proteins.
One member of the C19 peptidase family is ubiquitin specific protease 47 (USP47), a cysteine protease and deubiquitinating enzyme. While the sequence of USP47 has been identified, no molecular studies have been conducted. Since C19 peptidases are generally involved in regulating the cell cycle, characterizing the activities and roles of specific members would be useful for studying cell cycle regulation and in particular to study the effects on apoptosis.
There is a general need for cancer treatments that utilize methods and agents that regulate or affect the cell cycle to stimulate apoptosis of desired target cells, including cancer cells. Identification of the effects of inhibiting or blocking USP47 would be useful for studying cell cycle regulation and in particular to study the effects on apoptosis. In particular, there is a need for small molecule inhibitors of USP47 useful as cancer chemotherapeutics.